In the following article the writer has endeavored to briefly summarize a theory of the resistance of face-hardened armor and the action of projectiles, both plain and capped, on impact, with the further object of presenting perforation formulas derived from experiments at Indian Head.
Prior to the advent of face-hardened armor formulas for the penetration of wrought-iron and homogeneous steel plates were based on the assumption that the projectile experienced no change of form while passing into and through the target. With face-hardened armor, however, such a thing rarely happens under the conditions of test except when the plate is far outmatched by the projectile, as when the projectile is of extraordinary quality, or the plate decidedly inferior.
As the result of experience in the United States it is not believed that a good formula based upon anything but complete perforation can be constructed to express the relation between gun and plate on impact, when the latter is face-hardened. In homogeneous plates the penetration is always proportionate to the energy of impact, and the resistance offered is largely local, closely corresponding to the action of punching. The action of resistance is radically different with face-hardened armor. The inextensibility of the hard face and its inability to bend or flow transmits the strain of impact to the tough back, and distributes it over a considerable area, thus bringing to bear all the resisting power of that portion of the plate to check the advance of the projectile. If the projectile has sufficient remaining energy penetration ensues, otherwise the projectile smashes harmlessly on the plate. The first effect is to elastically dish the hard face; when the limit of strain is reached this gives way, and the resistance then becomes largely local; hard jagged fragments of the surface are carried into the body of the plate, scoring and abrading the ogival of the projectile and impeding its advance. The further function of the hard face is to prevent the flow to the front of displaced metal of the plastic body. Sometimes, when the projectile remains undeformed and the hard face is circumferentially broken to a diameter considerably greater than the caliber of the projectile, a front fringe appears around the shot hole; this of course decreases the total resistance to penetration.
It will thus be seen that the function of the hard face is threefold, of which the first mentioned is by far the most important. There is a definite limit of velocity corresponding to a considerable proportion of the total energy required for perforation, below which the projectile makes practically no impression on the plate, except perhaps a saucer-like depression enclosed by fine concentric cracks in the hard face. A 5-inch Carpenter projectile with a striking velocity of 1522 f. s. hardly marred the surface of a Carnegie 4-inch plate, while a velocity of 1977 f. s. just effected perforation. An 8-inch Holtzer projectile, striking a Bethlehem 10-inch plate at a velocity of 1498 f. s., failed to break the hard face.
The theory of the resistance of face-hardened armor, now generally accepted in this country, was first enunciated by Lieut. A. A. Ackerman, U. S. Navy, and his deductions have been fully borne out by subsequent experiments. It seems to be the popular impression that the resisting power of armor due to the action of the hard face is derived from its power to fracture the shot's point, with a subsequent smashing and pulverizing progressively from the point.
It is difficult to see how this theory can be reconciled to the results of the great number of experiments conducted in this country. On such an hypothesis disintegration of the projectile must commence the instant the point reaches the face, and the successive layers or particles of metal shearing and sliding over one another and the whole breaking into a myriad of fragments, thus distributing the remaining energy among the particles and rendering the aggregation powerless to do harm. When partial penetration is effected some of these fragments are supposed to act together as a unit, combining their energy—the whole mashing into the indent and forming a coherent mass that remains sticking in the plate.
On impact the following cases represent the action of the projectile depending upon the caliber of gun, thickness of plate, weight of projectile, quality of projectile and plate, and striking velocity:
1. Where complete perforation is effected, the projectile remaining intact or only slightly deformed. In nearly every instance where the projectile is upset, the diameter is increased around the bourrelet, and where there are signs of rupture cracks appear extending parallel to the axis emanating evidently from the upset portion, and forming ribbon-like strips evenly distributed around the circumference. 2. Sometimes a section of the plate is punched out to a diameter equal to or slightly exceeding the caliber of the projectile, the shell being broken up and part of the head remaining apparently welded into the face of the section punched out. (This is not usual in the United States where the plate is just perforated, though from the reports it seems to occur abroad, especially in Germany, in almost every instance where bare perforation ensues.) In this country the plate usually gives way in the rear by rupture along lines radiating from the center of the back bulge. Sometimes the punched-out section is found intact, and this usually occurs when the plate is much overmatched. The punched-out section is then nearly always in the form of an obtuse cone. 3. Where the head remains sticking in the plate. Sometimes, but rarely, the point and ogival remain undeformed, but usually the ogival is mashed into a mushroomlike mass, breaking away the shot-hole to a diameter sometimes 50 per cent larger than the caliber of the projectile; this is caused by its upsetting at or near the bourrelet. The apex of the powder chamber is invariably preserved when the penetration is appreciable. 4. Where the projectile rebounds, being shattered on impact, or remaining intact or slightly deformed. In the latter case the same effect is observed as when perforation occurs, namely, bulging at the bourrelet with a tendency to rupture along longitudinal lines extending through the bourrelet.
Often the point is broken or sheared off when the impact is not normal. It is rare that the axis of the projectile is normal to the tangent plane of the surface at the moment of impact. Even when striking at a considerable angle the projectile turns and enters the plate normally, following the line of least resistance. It is this lateral component of the blow at an angle to the axis that causes the breaking off of the point, the projectile often remaining otherwise undeformed after perforation or rebounding. The fracture in such cases is always clean, with no evidence of a sliding motion over one another of the particles, showing that in the case of perforation the point is carried through the plate, becoming detached afterward. Occasionally, when the projectile is of excellent quality, but the velocity unequal to perforation, the head of the projectile remains sticking in the plate apparently undeformed. A Krupp face-hardened plate exhibited at the World's Fair at Chicago in 1893 showed this strikingly. The back bulge on this plate was so cracked as to show the point plainly visible and apparently perfect in form.
The action of the projectile on impact with regard to rupture may be stated as follows: The first tendency is to bulge at or near the bourrelet; this is due to longitudinal compression and is what would be expected. At the moment of impact the greatest strain is on the point, but this does not give way locally, being in the first place superior in quality to the metal of the plate (and this latter would therefore give way first where the energy is sufficient to effect penetration at all), and in the second place being supported by the form of the ogive, which may be compared to an arch of which the point is the top. The strain is thus transmitted to the body of the projectile, which causes it first to bulge and then to crack longitudinally in response to the transverse component of the blow. When the head effects an entrance and the plate is not perforated, the hard envelope of the ogival is usually flaked off and the more plastic metal of the interior is moulded into a mushroom-like mass, not rupturing because of the support of the metal of the plate. Sometimes, however, the tempered surface of the ogive remains intact. This rarely occurs in partial penetration, but is nearly always the case when complete perforation ensues, all parts of the projectile getting through the plate. In the latter event when rupture occurs, the head breaks up along its weakest lines.
Finally, in all cases, fragments found either in front or in rear of the plate present clean fractures, except where the pieces are evidently marred by contact with the plate; the lines of fracture of the body are at right angles and parallel to the axis, and the exterior surface of the fragments (except in the cases cited before, where the hard envelope of the ogive is flaked off) shows that there was little or no deformation before rupture.
In effect the concentrated resistance presented by the face-hardened plate checks the advance of the projectile, and the sudden stoppage causes the latter to break up along its weakest lines through its own inertia.
As to the advantages of a soft steel cap fitted over the point of the projectile. Many experiments have been carried on by the Bureau of Ordnance with various forms of caps presented by Mr. Isaac G. Johnson and with some of its own devising, and the results have so uniformly demonstrated the superiority of projectiles so fitted in the penetration of face-hardened armor that the Department has decided to fit them to all service projectiles. A tabular list of all these experiments is appended to this article and reference will be made to this table further on. The cap, as adopted, consists of a cylindrical piece of soft steel, half the caliber of the projectile in diameter, bored out to a depth of two thirds its length to fit over the head of the ogival. A recess or cavity in the interior surface, .03 of an inch deep, contains a lubricating material. The sketch shows a service cap fitted to a six-inch projectile.
Several theories have been advanced to explain the action of this cap on impact. 1. Some hold that it acts as it were a buffer against the hard and impenetrable face of the plate, preventing the sudden stoppage of the projectile and greatly lessening the shock of impact. 2. The main claim of the patentee, Mr. Johnson, is in effect that the cap, completely surrounding and enclosing the point of the projectile, as it does, strengthens the projectile by supporting it all around, thus giving increased resistance to lateral deflection and to longitudinal compression.
It is believed that neither one of these theories offers the true explanation. The fallacy of the first becomes apparent when it is considered that if the function of the cap is merely to lessen the shock of impact, the same results could be achieved by firing a projectile without a cap at such a reduced velocity that it will strike the plate at the same velocity as the capped projectile had after piercing the cap. In other words, unless the cap in some manner weakens the plate or strengthens the projectile it can only reduce the velocity with which the point of the projectile meets the face of the plate.
2. It is difficult to see how the hard and superior metal of the point of the projectile can derive any considerable support from the soft metal of the cap, offering as it does but an insignificant resistance to longitudinal and transverse strains. Certainly it cannot resist materially the compressive strains due to impact. Possibly the cap may be a slight aid to the point in preventing it from being broken off when the impact is oblique, but it is believed that the cap is disintegrated in this case before the shearing strain comes on the point. This theory was no doubt derived from the supposition that projectiles crush progressively from the point. If the cap is ruptured, as would seem to be the case, before the point of the projectile reaches the face of the plate, the cap would operate merely to reduce the striking velocity.
The true explanation of the action of the cap must be taken in conjunction with the theory of the resistance of face-hardened armor enunciated before. The cap, meeting the face of the plate at a high velocity, dishes or depresses the hard surface to the limit of its elasticity, the cap being itself destroyed in so doing. And to be effective, therefore, the cap must have sufficient stiffness or longitudinal strength to accomplish this; possibly the hard face is crushed in by this action. The projectile meanwhile is advancing, its way through the cap being smoothed by the lubricant, and when the point meets the plate resistance to its advance is purely local. Possibly the passage of the projectile through the plate is facilitated by the carrying in of sufficient of the lubricant and portions of the soft cap to cover asperities in the metal and prevent the hard jagged fragments of the face from scoring and abrading the surface of the ogival. This, it is believed, is the true function of the cap. The comparative ease with which face-hardened armor is perforated when a thin plate of wrought iron is placed over the surface, as shown by Russian experiments, can be readily explained on this theory, by considering the wrought-iron metal opposed to the point of the projectile at the moment of impact to bear the same relation as the cap to the point of the projectile. The action is explained in the following somewhat exaggerated sketch (see next page):
The point of the projectile enters the wrought-iron plate and the stress of impact is transmitted to the hard face along lines normal to the ogival, depressing the surface to the limit of its elasticity, so that when the point reaches the plate, although its velocity is reduced, the resistance it encounters is not concentrated but is local. The breaking in of the hard face by the superior metal on the point of the projectile is then comparatively an easy matter.
Various forms of these caps have been tried at Indian Head with more or less definite results. The following, however, seem to be conclusively established:
1. That a projectile fitted with a solid cap of the form finally adopted, but containing no lubricant, is superior to an uncapped projectile.
2. That a cap in the form of a hollow cylinder with thick walls, containing no lubricant, is equally as efficient as the solid cap containing no lubricant. This goes far toward proving the theory that the effect of the cap is to weaken the plate rather than strengthen the projectile.
3. That a thin-walled envelope filled with lubricant does not facilitate penetration.
4. That the most effective form is a thick-walled cap or envelope, strong enough to withstand considerable strain before rupture, yet plastic enough to permit of some deformation before breaking, combined with a lubricant contained in a recess or cavity surrounding the point of the projectile. Soft steel seems to be the best material for the cap, though experiments have shown copper to be very efficient.
In the London Engineer of July 3rd, 1896, there is an article entitled, "Herr Krupp on the Perforation of Steel Armor," in which is mentioned a formula recently propounded by Krupp as applicable to the perforation of the best and newest plates with hardened faces. The original formula given in continental units is p v2= 5800 a e, where v is the striking velocity in meters, p the weight of the projectile in kilos, a the diameter of the projectile in cm., and e the thickness of the plate in cm.
According to Krupp, continues the writer, we may now take it as based on the best formula that can at present be suggested. And being founded on correct principles, the formula will no doubt be fairly accurate if the constant is changed to meet changed conditions so long as the projectile is assumed to pass into and through the target unbroken and undeformed.
Experiments ranging over a period of several years, comprising the attack of face-hardened armor by guns of all service calibers up to and including the 13-inch, have been carried on by the Bureau of Ordnance, under the direction of Captain W. T. Sampson, U. S. Navy, at the Naval Ordnance Proving Ground at Indian Head. The similarity of conditions in these experiments, in quality of armor and projectile, of gun and charge, furnishes reliable data from which a formula can be evolved that will express with fair accuracy the relation between the perforating power of the projectile and the resisting power of the plate.
A table is appended giving the details of those experiments that were chosen as being of a thoroughly representative character. In a large majority of these cases where perforation ensues the energy of the projectile was just sufficient to overcome the resisting power of the plate, completing perforation with but slight surplus energy, the plate cracked possibly and the projectile set up, deformed or disintegrated, the pieces lodging in the backing or in the butt. In a great many cases reservation must be made due to the weakening of the plate from previous impacts.
It might here be stated that where plate and projectile are practically evenly matched, being of good quality, both are strained nearly to the limit of rupture, and when this occurs either in one or the other it does not mean a considerable increase in the consumption of energy beyond that taken up by the strains of impact. Assuming, then, an equation of the form, it will readily be seen that such indeterminate elements as the cracking of the plate, breaking up of the projectile, dissipation of energy in the form of heat, can, by considering the absorption of energy in these ways to bear the same proportion in all cases to the total striking energy, be included in the constant. This equation is fundamentally sound. It was first elucidated by Fairbairn on the hypothesis that the destructive power of a shot was due to its energy, and has since formed the basis of all formulas for the penetration or perforation of armor.
Having, then, data deriving considerable reliability from similarity of conditions, it becomes easy to evolve a formula that will express within limits, allowing for difference in quality of plate and projectile, the results of the laws of destruction, though of course such an expression being empirical, the laws themselves cannot be definitely stated.
As before stated this formula expresses the relations between the elements only when perforation ensues. From it can be determined the minimum velocity sufficient to perforate a plate of given thickness, or the maximum thickness of plate that can be perforated by a projectile of given velocity. The non-deformation of the projectile is not assumed, but the formula is based on data where the energy of the projectile was barely sufficient to accomplish perforation, the projectile being broken up and the pieces passing through the plate.
In Germany there are records of many experiments against plates of different thicknesses, which, while not so extensive as those made in the United States, are still sufficient to give a fair idea of comparative value. There is no question but that Krupp armor is at least the equal of if not superior to that of England, France, Austria and Russia. Experiment indicates the latter; hence the other countries will not suffer if Germany is taken as the European representative in comparing the quality of American armor with that made abroad.
Some of Krupp's most recent experiments are tabulated and appended with columns showing the velocities required for perforation under the same conditions calculated from the proposed formula and from the one quoted before as being that presented or used by Krupp. The agreement in the former case with the actual results is striking. The velocities given by the Krupp formula seem to be entirely at variance with the actual results for the thicker plates; too much value is given to slight variations in thickness of plate. Captain Castner, however, states in an article published in "Stahl und Eisen" in April, 1896, referring to the Krupp formula: "Whether this formula can also be applied to thick plates is, according to the assertion of the Krupp works, still unsettled, as such plates have not yet been perforated."
These experiments of Krupp with the larger plates were evidently conducted with the view of obtaining the actual velocities required for the perforation of the plates. The truly remarkable agreement of the calculated with the actual velocities as given by these experiments shows that much reliability can he given to the proposed formula as a basis for comparing the resisting qualities of German and American armor. An examination of the tabulated results of these experiments at Indian Head, upon which the formula is based, carried on during a period of several years, shows a constant and steady improvement in the quality of Carnegie and Bethlehem armor. Attention is called to the Bethlehem 8-inch plate, manufactured for the Maine's turret and tested October 2, 1894. As regards resistance, the best service plate yet tested was a Carnegie 12 ½-inch reforged, face-hardened plate, B. 491, manufactured for the forward 13-inch B. L. R. barbette of the Kearsarge.
The plate had previously been attacked by two 10-inch Holtzer A. P. shell, the second of which made a diagonal crack 5 feet long and 7 inches deep in the back of the plate. This plate was attacked by a 12-inch Carpenter projectile weighing 85o lbs. with a striking velocity of 1932 f. s. and a striking energy of 21,940 tons. The back bulge was driven out, but no part of the projectile, which broke into 79 pieces, got through the plate. This velocity under similar conditions would, by the formula, have sufficed for the perforation of a 14.3-inch plate. The projectile, as stated by the Inspector of Ordnance, gave every indication of being an excellent one. The plate was through cracked and broken into three pieces, lines of fracture radiating from this impact and passing through place of impact No. 2.
By the formula, using the weight of a Krupp or St. Chamond 12- inch A. P. projectile, 716 lbs., a plate of the thickness of 14.3 inches requires a velocity of 2107 f. s. to perforate, which, in the experiments on plate 575-A, was shown to be just sufficient for the perforation of a plate 13.78 inches thick. This latter plate seems to have been more resisting than its companion 14.49-inch plate, 575-B, and slightly superior to the average of American plates as shown by the formula. Therefore the 12.5-inch Carnegie plate would seem to be equal in resistance to a 13.78-inch Krupp plate. There are several considerations, however, that must be taken into account affecting the final comparison of superiority. The Krupp plate was experimental and no doubt manufactured with the utmost care with the view of obtaining the most resisting plate; the Carnegie plate was chosen from a lot of twenty-two as promising to make the poorest showing ballistically of all. On the other hand, the Carnegie plate was practically destroyed, being cracked and broken into three pieces, while the Krupp plate was not through cracked, being indeed hardly cracked at all; such cracks as did develop being confined practically to the surface, and radiating from the point of impact to the edges. The greatest depth of these cracks was about two inches. Again, the greater weight of the Carpenter projectile is no inconsiderable factor even though the energy of impact in both cases is the same.
Weighing these considerations, then, it would seem that the Carnegie and Krupp plates were of about equal merit, the superior resisting power of the one being balanced by the greater toughness and tenacity of the other.
The most resisting plate that Krupp has yet produced is the H.8-inch plate tested at Meppen in September, 1895. This plate withstood the impact of a Krupp 712.6 lb. A. P. projectile with a striking velocity of 1993 f. s., being practically uninjured save three surface cracks, the greatest depth of which was 3.1 inches. The plate was not perforated; the depth of penetration was not given, but from the fact that the back bulge was 3 inches high and slightly cracked it would appear that the limit of resistance had been almost reached. A Carnegie 12-inch experimental reforged face-hardened plate, tested at Indian Head on May 29, 1897, compares very favorably with this star plate of Krupp. The Carnegie plate was 12 feet long, 8 feet wide and 12 inches thick, and was backed by 12 inches oak and two %-inch skin plates, the backing being secured to the plate by 18 2.8-inch armor bolts. In shape the plate was flat, rectangular. It was attacked first by a 12-inch Holtzer A. P. projectile weighing 85o lbs., striking with a velocity of 1811 f. s. The impact was 6 feet 2 inches from the right edge and 3 feet 8 inches from the bottom; the point of the projectile just perforated the back bulge, punching out an attenuated hollow cylindrical section that was driven through the backing and fell in rear of the plate. The surface of the plate was dished ,9-inch over an area corresponding to a diameter of about 3.5 feet; diameter of flaking, 21 inches; diameter of shot-hole, 135/2 inches, the interior being quite rough. The plate showed no signs of cracking. The projectile broke up, some of the pieces getting through, but the bulk falling in front of the plate. This projectile appeared to be of very good quality. It was evident that the plate and projectile were nearly evenly matched, and that with a slightly decreased velocity the projectile would have been defeated. A second round was then fired, using an 850-lb. Wheeler-Sterling projectile, striking velocity, 1769 f. s.; striking energy, 18440 f. s.; location of impact, 3 feet from left edge and 3 feet from the bottom. The projectile smashed on the plate, a portion of the head remaining stuck in the impact. The surface of the plate around the impact was dished and flaked as usual. There were no signs of cracking.
Now, by the formula, the Krupp plate should be perforated by a 12-inch 712.6-lb. projectile with a velocity of 1829 f. s. It with stood, however, a velocity of 1993 f. s. The Carnegie plate by the formula should be perforated by a 12-inch 850-lb. projectile with a velocity of 1696 f. s.; it was just defeated by a velocity of 1811 f. s. On the face of it, the Carnegie plate, to have been equal to the Krupp plate (the relative quality of the projectiles not being considered), should have defeated an 850-lb. projectile at 1846 f. s. It must be remembered, however, that the angle of impact in the instance of the Krupp plate was from the normal, while in the case of the Carnegie plate the impact was exactly normal; the former plate was slightly cracked after three impacts, while the latter showed no signs of cracking after two rounds. All in all, it may fairly be said that this Carnegie plate is fully as good as that of Krupp.
As to the relative value of the qualities, hardness and toughness, opinions differ. In the development of the art of manufacture of face-hardened armor Krupp seems to have taken the maximum of the latter as the objective point with the greatest amount of hardness consistent therewith; while in this country resistance to penetration has been the prime object. The writer is of the opinion that a degree of hardness could well be sacrificed to increase toughness. The armor makers, however, having reached such a high standard of resistance, are pardonably loath to give up even a modicum of it, and are now striving to increase the toughness and still retain the excellence of resisting power to which they have attained, with encouraging hopes of success, as shown by the experiment described above.
The foregoing remarks apply to the thicker plates; the thinner plates, German and American, seem to possess similar qualities of resistance, both as regards hardness and toughness, with a slight advantage, if any, in favor of the American plates.
The adoption of soft steel caps fitted over the points of projectile has taken away at least 15 per cent from the efficiency of face-hardened armor. The following formula, tentative through lack of sufficient data, especially for the larger calibers of gun and plate, to support its reliability, is proposed for the perforation of face-hardened armor by capped projectiles.
The tremendous advantages of the higher caliber guns and heavier projectiles is apparent. The sustained velocity and destructive effect due to the weight of the projectile are factors of prime importance. No armor has yet been made that can withstand the terrible blow of a 13-inch shell, fired with a muzzle velocity of 2400 f. s. (which will hereafter be the service velocity, using smokeless powder), and striking normally within a range of 2500 yards. The value of the 8-inch gun may be clearly seen, and arguments for its retention in future battle-ships should be convincing.
The tendency abroad has been to decrease the thickness of armor, the caliber of guns and the weight of projectiles for the different calibers, a curious reaction from the policy in vogue not many years ago when equality between gun and armor was maintained by increasing the power of the former and the thickness of the latter, the quality of both armor and projectile undergoing steady improvement.
The adoption of the face-hardened process almost simultaneously, in this country at least, with the use of nickel in the steel, has resulted to the present time in an increase in efficiency over homogeneous steel armor of 8o per cent for thin plates, 4 or 5 inches in thickness, ranging to about 45 per cent for the very thick plates, and an increased efficiency over oil-tempered nickel steel armor of at least 35 per cent on an average. The advent of the cap has, however, reduced this lead to 20 per cent. The importance of the latter, therefore, must be appreciated, especially as caps can be cheaply fitted to the armor-piercing projectiles now in service.
The logical projectile so far as the power of perforation goes is undoubtedly the solid shot, but its advantage in this regard over shell is more than outweighed by the destructive effect of the latter, due to its breaking up after perforation, especially as armor-piercing shell can now readily be burst after piercing a considerable relative thickness of armor, equal, say, to the caliber of the projectile.